Research On Nondestructive Testing Technology Of Solar Cell Based On Image Processing

As a carrier of photoelectric conversion,solar cell quality directly affects the power generation efficiency and service life of the product.the defect detection technologies of visible light imaging images and electroluminescence images of solar cells were studied with the image enhancement and image segmentation techniques of digital image processing.First,the solar cell image surface defect detection algorithm is studied.In order to obtain high-quality solar cell image,the LED strip-shaped combined light source is used to simulate natural light,and an image acquisition platform is built and the solar cell image is collected.The collected image is preprocessed by means of median filtering and morphological processing to reduce interference information in the image.Next,surface defects such as scratches,falling pieces,and dirt of the battery sheet are extracted by the Otsu algorithm-After that,the chamfering and electrode regions of the solar cell sheet after the Otsu algorithm treatment were filled and the defect rate of the solar cell sheet was calculated.Compared with the related algorithm,the defect area detected by the algorithm is closer to the actual area of the defect,and the detection accuracy is about 2 times that of other algorithms.Secondly,the contrast between the defect and the background in the EL image of the solar cell is small.In order to improve the contrast of the image,an improved double histogram equalization algorithm is proposed,which can improve the image contrast while widening the image contrast.At the same time,an image segmentation algorithm based on global threshold and Frangi filtering is proposed,which can accurately extract the EL image defects of solar cells.The data shows that the accuracy of defect detection in this algorithm is nearly 4 times that of other algorithms.Finally,the characteristic features such as geometric features and gray features are used to describe the detected defects,and the characteristics of the characteristic parameters of different defects are analyzed and summarized to help the defect classification and identification.